Gathered reports on the accuracy testing for the Forest Disturbance Service using bfast and the harmonized Sentinel-2/Landsat-8 collection. The different reports are showing: - event / no-event tests - different length of the stable history (for no events) - different parametrization
1
2
3